Psoriasis linkage to 4q28-32 (PSORS9) was initially identified by our genome-wide scan in 61 Chinese families and subsequently supported by a meta-analysis of five genome-wide linkage scans of European populations. In this study, we performed a follow-up analysis of PSORS9 using an additional 90 families and improved marker coverage. Joint analysis of all 151 families obtained significant linkage evidence (HLOD=4.53, nonparametric linkage (NPL)=4.03 (P=0.000003)) at the marker interval D4S2997-D4S3033, and the same was obtained for the analysis of the independent new families (HLOD=4.33, NPL=3.15 (P=0.00004)). The linkage evidences from the whole families and the new families exceeded the genome-wide criteria for significant linkage. Furthermore, by performing an ordered subset analysis using mean age at onset as a covariate, we demonstrated that evidence for linkage to PSORS9 is concentrated in the early-onset families and suggested that further study of PSORS9 should focus on early-onset patients. This finding is contradictory to what was found in the Icelandic population and, together with other linkage results, suggests that Chinese and European populations are genetically different for linkage to PSORS9, which may partially explain the influence of ethnic factors on the varying prevalence of psoriasis.